READOC: A Unified Benchmark for Realistic and Comprehensive Document Structured Extraction
READOC is a novel benchmark that frames document structured extraction as a realistic, end-to-end task of converting unstructured PDFs into semantically rich Markdown text, enabling a unified evaluation of state-of-the-art approaches.
多文書を一括処理可能な座標認識付き文書解析モデル CREPE
CREPEは、単一の文書画像に複数の文書が含まれる場合でも、各文書の解析結果を区別して出力できる。また、文書内の文字列の位置座標も同時に出力することができる。
A Comprehensive Dataset for Multimodal Information Retrieval in PDF-based Visual Question Answering
PDF-MVQA is a new dataset that enables the examination of semantically hierarchical layout structures in text-dominant documents, allowing the development of innovative models capable of navigating and interpreting real-world documents at a multi-page or entire document level.
LayoutLLM: Enhancing Large Language Models for Document Understanding through Layout Instruction Tuning
LayoutLLM is an LLM/MLLM based method that integrates a document pre-trained model as encoder and employs a novel layout instruction tuning strategy to enhance the comprehension and utilization of document layouts for improved zero-shot document understanding.
Unified Framework for Visually-Situated Text Parsing: Text Spotting, Key Information Extraction, and Table Recognition
OMNIPARSER is a unified framework that can simultaneously handle text spotting, key information extraction, and table recognition tasks through a single, concise model design.
M3-VRD: Multimodal Multi-task Multi-teacher Visually-Rich Form Document Understanding
다중 교사를 활용한 복합적인 문서 이해 모델의 효과적인 성능과 지식 전달
mPLUG-DocOwl 1.5: Unified Structure Learning for OCR-free Document Understanding
統一構造学習によるOCRフリー文書理解の向上
GRAM: Global Reasoning for Multi-Page VQA
提案されたGRAMアプローチは、既存の単一ページ文書モデルを拡張し、多ページドキュメントを効率的に処理する方法を提供します。
Unified Structure Learning for OCR-free Document Understanding with DocOwl 1.5
Enhancing Multimodal Large Language Models with Unified Structure Learning for improved OCR-free Document Understanding.
Global Reasoning for Multi-Page VQA: GRAM Methodology
Efficiently extending single-page models to handle multi-page documents in visual question answering.
